Mike Silva teaches us a neat lesson

In a post entitled, “wRC+ = Follow the Money Trail,” Mike Silva of NYBD figures out the dastardly plans of the online saber community: make that money. Take it away, Mike:

We need more advanced statistics just like we needed Cool Ranch Doritos, New Coke, or a colorful cover for the iPod. What I mean is the “powers that be” in the statistical community have created a profitable industry for themselves that results in sites like Fangraphs. Just like newspapers churn out content daily, statistical analysis becomes stale if it isn’t procured with another variation of the same thing. Why have boring old Runs Created, or OPS+, when you can fancy it up with wRC+.

Tango (seen here) responded to Silva in the comments section by saying:

As I keep saying on my blog time and time again, summary opinion without evidence is the very definition of bullsh!t. And Mike’s statement here is pure bullsh!t.

And the funny thing is how right Tango is. You see, the neat lesson Silva has taught us today is the beauty of meritocracy in the online saber community. It would be near impossible for Tango or Appleman to bring out the same old stat spiced up with a different name and get away with it. Why? Because there are smart people here on the internet who would never put up with being treated like idiots. These aren’t mindless hoards of consumers begging for Michael Bay to blow things up for the billionth time. These are intelligent fans, some of them specialists in mathematical and analytical fields, who demand good evidence be given at all times. If Dave Appleman were to go on Fangraphs and say, “Hey guys, we have a new stat: wUZROBA+, and it’s made by doing this nonsense, etc…” people would stop reading Fangraphs! Instead, he says this:

As you may have noticed, there’s now an extra column in the “Advanced” section for batting stats called “wRC+”. You can think of this stat as a wOBA based version of OPS+. It’s park and league adjusted and it’s on a very similar scale as OPS+. The difference is that it uses wRC, which is based on wOBA.

There’s the evidence. Now whether or not this is a good stat or not can be discussed. How? By actually analyzing the statistic based on its inherent merits, which Silva never does! Mike responds by saying:

You don’t need to go to an Ivy League school, create a metric, or understand science to have intelligence (looking at you Andy), perhaps if some of the people skills that, from my experience, many in the sports industry lack, were present you would be able to sell your concepts to the generally public rather than arrogantly responding (looking at you MGL) your critics.

And I agree, you don’t need to have a PhD (or any degree) to be a knowledgeable baseball analyst. But you do need to demonstrate tangible proof for the claims that you make, and this is where Mike fails greatly. In fact, he admits it:

Front offices all across the country are riddled with politics, am I naïve enough to believe the advanced metric community is no different? Do I have proof? Of course not, I can’t read your mind… You will learn that many things in this world at not x +y = z…All I know is new stats continue to be churned out and sometimes I really don’t see where they are nothing more than the proverbial rocking chair.

Here’s how you can uncover Tango, MGL, Appleman, and the rest, Mr. Silva: demonstrate with evidence that their stats are nonsense, and you win. Seriously. Just prove that wRC+ is utter and complete dribble and you can advance your position a lot. But Silva, admittedly, doesn’t have evidence. He has a hunch. A gut feeling. He sees analogies to other entities (that aren’t analogous) and creates links. And this, at its heart, is what is separating the saber community from the rest of the pack.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: The most underrated catchers in the game
Next: Playing chicken with Boras »


  1. James K. said...

    Hopefully no one actually clicks through to the original post.  Nice writeup Pat. Regardless of Silva’s nonsense you made a lot of great points about sabermetrics and the saber community.  Hopefully this is the last time Silva garners any attention from the rational crowd.

  2. Dave Studeman said...

    Got to say, I agree with Mike Silva’s general point (though I haven’t read his original post).  There’s nothing “wrong” with the new stats, but the extra insight they yield is marginal.  At the same time, they run the risk of making sabermetrics unapproachable to the majority of fans.

    Every “generation” of baseball analysts feels compelled to create its own stats.  That’s always been true, and it will always be true. That’s the fun of it. And the sites that display those stats will be the favorite sites among the active sabermetricians of the day.  At THT, we refrained from creating too many new stats and we lost a lot of traffic to Fangraphs. You create “sizzle” and traffic by creating and displaying new stats.

    Fangraphs, btw, is a great site and I’m proud to be associated with David and his crew (same with Baseball Reference).  But that is the reality of how sabermetric fandom works, IMO.

    At the same time, every generation will defend their stats, and they will almost certainly be correct in doing so.  That doesn’t undermine the point, however. Creating new stats and using them to draw traffic to your site is human nature.

    BTW, if anyone believes that Fangraphs or THT is turning anything resembling what a businessperson would call a “profit,” they are sorely misinformed.

  3. Joe R said...

    I agree with what Dave said. wRC+ is a moderate upgrade to OPS+ and wRC on their own, but for a casual fan, throwing that statistic around at them could turn them off to good analysis in general. Most fans I know, despite everything, still start their baseball related prayers with “In the name of the batting average, and of the home run, and of the holy RBI”, and telling them (before this year at least) that Jack Cust’s bat is more productive than Alfonso Soriano’s, or that Joe Carter was one of the worst players ever with his career length (well maybe they would buy that).

  4. Dave Studeman said...

    So we should ignore/avoid marginal upgrades in the fear that their reception would be negative?

    No, but you should at least respect the opinions of those who see no use for the marginal upgrades, and you should also recognize that you may be setting back the “cause” of getting sabermetric ideas accepted by more baseball fans.

    Of course, a lot of sabermetricians consider that a lost cause anyway.

  5. dkappelman said...

    Dave,  I completely agree with you that there’s no point in adding stats just because they exist or are new.

    I think over the past year there’s been very few new stats added to FanGraphs.  wRC+, xFIP, tRA, and the pitch values are the only actual stats that have been added to the site and there’s not a whole lot of plans to add additional raw stats, though there are some things I’d like to see such as baserunning and catcher defense.

    I believe, FanGraphs has a mostly complete statistical toolbox on how to evaluate a player and unless there’s a stat that brings something new to the table or makes stats in general more accessible, then I’m really not interested in adding it.  I happen to think wRC+ fell into the accessibility category.  I’d rather fine tune existing stats than create an ever increasing library of them.

  6. Funes said...

    Every “generation” of baseball analysts feels compelled to create its own stats.  That’s always been true, and it will always be true. That’s the fun of it. And the sites that display those stats will be the favorite sites among the active sabermetricians of the day.

    Part of what is disappointing about this cycle is the way that “new stats” are often integrated into the existing body of knowledge.

    Often, not enough attention or explanation is given to the underlying construct the new metric is supposed to represent.  This can create confusion and fragmentation between “generations” rather than the organic growth of one set of ideas out of another.

    Just as an example – take wRC+ and how it was explained:  “You can think of this stat as a wOBA based version of OPS+.”  That’s not an explanation of the idea that wRC+ is supposed to represent.  And what makes this lack of conceptual clarity even greater in this case is that wRC+ is based on wRAA – not on wOBA. From this description I don’t see any tangible, concrete detail for how wRC+ is in fact conceptually related to either wOBA or OPS+.  Having the same scale isn’t an explanation.

    So now there are people all over the blogosphere quoting wRC+ and it seems like many of them don’t even know what the underlying concept the metric is supposed to represent is and how it is related to the existing body of sabermetric knowledge.  It opens up the new stats as seeming capricious and gimmicky.

  7. Tom M. Tango said...

    “No, but you should at least respect the opinions of those who see no use for the marginal upgrades,”

    Right, I respect those opinions by ignoring those opinions, just as much as they ignore my opinions and blog.  Ignoring each other is perfectly fine.

    The problem is when you go from ignoring to DISMISSING.  Where “ignoring” means not to go to someone’s party, “dismissing” means to tell other people that that party sucks.

    If they were to tell people WHY the party sucks, then that would be fine.  He provides value.

    But, Mike’s article was all about dismissing things he did not understand.

  8. Tom M. Tango said...

    “wRC+ is based on wRAA – not on wOBA”

    Actually, all of these stats are based on Pete Palmer’s Linear Weights.

  9. Funes said...

    Actually, all of these stats are based on Pete Palmer’s Linear Weights.

    Sure.  I know.  And that’s my point.  If all of these “new metrics” were explained in a fashion which made them grow out of a common body of growing knowledge that has a clearly defined conceptual basis that would help a great deal.

    So explaining wRC+ as an extension of linear weights would be much clearer than trying to contort it to a wOBA version of OPS+.

    And following the link from the Fangraphs page to The Book, much of the discussion through the thread is how wRC+ is most directly an extension of wRAA.  So while it is true that all of the stats are an extension of linear weights – that’s not the context for how wRC+ was introduced or explained either on Fangraphs or on the reference provided.

  10. dkappelman said...

    Funes, if you think it was poorly explained on FanGraphs you can put that on me.  My reasoning for presenting it the way I did was that most everyone on FanGraphs is familiar with wOBA and there was a lot of discussion about why wOBA+ wasn’t used, so I framed it in such a way where it was an extension of wOBA. 

    And then you wanted to learn more so you clicked on the link which gave you all the details, which you had obviously clicked on.

    I’m not sure though that if I had explained it some other way this whole discussion could have been avoided….

  11. Tom M. Tango said...

    If all of these “new metrics” were explained in a fashion which made them grow out of a common body of growing knowledge that has a clearly defined conceptual basis that would help a great deal.

    Well, I think I do explain them like that.  The Book is all about the “run expectancy matrix”, and how everything, Linear Weights (which is wRAA on Fangraphs), wOBA, Win expectancy, Leverage Index spins off that. 

    As for wRC+, David linked to my blog in his intro post:

    “Big thanks goes out to Tangotiger for pointing out how easy it was to implement this particular stat.”

    Which links to here:

    wRAA is park-adjusted?  I didn’t remember that.  Well, then, it’s a snap to convert to wRC+:

    a = league runs per PA
    b = wRAA/PA
    c = a/b + 1
    d = 100*c

    So, suppose someone is +30 (adjusted) runs, with 600 PA.  And the league average is .12 runs scored per PA (say 4.56 runs on 38 PA).

    a = .12
    b = 30/600 = .05
    c = .05/.12 + 1 = 1.417
    d = 142

    ta-da!  easy right?

    If you want to argue that Mike Silva did a bad job of saying what wRC+ was, fine.  If you want to argue that David should have cut/pasted, rather than linked, to my explanation above, I guess that’s ok too.  He did, later in the comments of that thread, cut/paste the relevant passage.

    Regardless, anyone who visits my blog knows where I’m coming from with all this stuff.  If you want to argue that it seems that you are walking into the middle of a conversation, then yes, you are right.  And that maybe we’re speaking Italian, and you only speak French and Spanish (close enough that you might be able to follow along), then yes, guilty.

    That’s why this site has some value:

    But, it takes time.

  12. Dave Studeman said...

    Hey David, my comments were meant to be general, not specific to this situation.  I can see how wRC+ fills in a gap in Fangraphs’ stats. Makes a lot of sense to me.  You know I think you do a fantastic job.

    Tom, I understand the point about ignoring vs. dismissing.  I assume you’re referring to some part of Mike Silva’s post that wasn’t included in Pat’s copy and paste, though his follow-up comments are ridiculous.

  13. Joe R said...

    My favorite part of Silva’s “argument” (more like trollbaiting) was saying that wRC+ doesn’t provide a better understanding of things.

    Take Mike Cameron (.346 wOBA) and Dexter Fowler (.345 wOBA). On paper, that’s the same.

    Cameron wRC+: 113
    Fowler wRC+: 103

    Suddenly the gap between their offensive seasons is far greater. So yeah, looks like wRC+ does provide plenty of understanding.

    (Brian Giles’ career wRC+ is 140, Mark Teixeira’s is 138, btw. Suck on that, ESPN).

  14. Funes said...

    I’m not sure though that if I had explained it some other way this whole discussion could have been avoided….

    Fair point.  I was trying to make a more general point and that was simply one example that came to mind. 

    I think the wRC+ data is useful and a nice addition.

    And it very well might be true that however you slice it, it’s difficult.  But my general sense is that drawing more of a connection between the similar ideas underpinning these individual metrics in isolation might help bridge some of the gaps in “generation” Dave Studeman was speaking to earlier.

    Many of these newer metrics are different views of similar data (e.g. wRC+ vs. wOBA vs. wRAA) and each provides different information. 

    So I didn’t mean to particularly focus on the explanation of wRC+ in and of itself.  I was more trying to say that if there was more conceptual context antecedent to the development of this particular new stat then it could be explained in a more direct fashion then by analogy to wOBA.  But perhaps that’s not the case.

  15. Mike Fast said...

    Silva’s assertion that Tango is a selfish fraud and nefarious huckster pretty much destroyed any facade that he was just concerned about clarity of discussion or the best interests of the game.  Although I respect Studes a great deal, I have to disagree with him 100% here.  Silva can take the insults and BS he flung toward Tango and MGL and turn them on himself. They’ve both given a lot to the game and have been as open to criticism and inspection of their methods as anyone.  It’s ridiculous for Silva to continue to whine they need watching and someone like him to cool their heels.  Guys like Tango, MGL, Studes, Alan Nathan, etc. are a big reason I’ve been able to do what I do and why I enjoy baseball analysis so much. They allow everything they do to be picked apart by the masses, and they’ve given a hand up to many others. It’s too bad that some folks out there mistake human kindness and gratitude for groupthink and an Illuminati plot to take over our beloved national pastime.

  16. Dave Studeman said...

    Mike, I don’t disagree with anything you said, let alone 100%.  I haven’t read Silva’s original post (and don’t intend to). But I do think the broader point, which I’ve tried to articulate in my own way, has merit.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>